Think about the next version of Scratch. What would it be if you could have a Voice Recognition block in your game? It would be like this: If players have a Mic, They can use Voice Recognition to make games do things. You can go to your "Voice" bar and select a block from voice. It is the same as any other, You click, Then drag form a multiple selection of Blocks. Then, When a player comes to play your game, You can use voice recognition to do it! All you have to do is set which word you want it to understand, Then test it!
And, We can even add in Text-To-Speech!
The computers can say what the creator inputs it to say.
-Voice Recognition-
___________________________
|When [ ] is said, [ ]|
-----------------------------------------
___________________________
|After [ ] is said, [ ]|
----------------------------------------
_____________________
|If [ ] is said, [ ]|
--------------------------------
-Speech-
___________
|Say [ ]|
-----------------
_______________________________
|After saying [ ], Say [ ]|
-----------------------------------------------
We must do this! It will work, And sometime in the future, I CAN SEE THEM INVESTING IN THAT!
Last edited by Richard2000 (2010-06-23 18:21:23)
Offline
http://scratch.mit.edu/forums/viewtopic.php?pid=331414#req_message
SOUND
Offline
Speech recognition is still pretty horrible, so having it in Scratch is just crazy.
Offline
You do realize that speech recognition that works is currently impossible for a normal computer?
Offline
juststickman wrote:
You do realize that speech recognition that works is currently impossible for a normal computer?
Actually, It's possible on a Macintosh/Mac.
Offline
Richard2000 wrote:
juststickman wrote:
You do realize that speech recognition that works is currently impossible for a normal computer?
Actually, It's possible on a Macintosh/Mac.
And on Windows.
Offline
gershmer wrote:
Richard2000 wrote:
juststickman wrote:
You do realize that speech recognition that works is currently impossible for a normal computer?
Actually, It's possible on a Macintosh/Mac.
And on Windows.
Only Windows Vista, Windows 7, and Macs have speech recognition.
Offline
...and My Windows XP laptop.
Offline
You can use volume spikes to determine syllables.
<when green flag clicked>
<set{ volume spikes }to(0
<forever>
<wait until><loud?>
<wait until><not> <loud?> >>
<change{ volume spikes }by(1
<end>
[/blocks]
For Speech Synthesis, you can record small fragments of speech and play them back in order - see my "Scratch Mail" project for an example.
EDIT: You can also use volume spikes for other things, such as claps, people in the room, music video cues or other audio events.
Last edited by bigreader (2010-03-02 14:41:16)
Offline
Speech recognition is not impossible, but it is cumbersome and buggy.
Unless there is a good squeak library with those functions in it, I cannot see the scratch team investing their time in this.
Offline
ThePCKid wrote:
gershmer wrote:
Richard2000 wrote:
Actually, It's possible on a Macintosh/Mac.And on Windows.
Only Windows Vista, Windows 7, and Macs have speech recognition.
All windows computers from 2000 to 7 have Text To Speech and Voice Recognition.
All Macintosh computers from System 7 to 10 have speech recognition.
It is possible to put in voice recognition for Scratch 1.5, We just have to work hard on it. And putting text to speech on the next version is completely a "Can-Do" thing.
If we use voices that are installed on Macintosh and Windows systems, Then it will be easy.
All we have to do is make it use the Voices that are installed on the system.
For first timers, It will need to fish for the fonts (Find them wherever they are) and load them onto Scratch.app or Scratch.exe.
Offline
bigreader wrote:
You can use volume spikes to determine syllables.
<when green flag clicked>
<set{ volume spikes }to(0
<forever>
<wait until><loud?>
<wait until><not> <loud?> >>
<change{ volume spikes }by(1
<end>
[/blocks]
For Speech Synthesis, you can record small fragments of speech and play them back in order - see my "Scratch Mail" project for an example.
EDIT: You can also use volume spikes for other things, such as claps, people in the room, music video cues or other audio events.
I guess that would work well, But it would be harder to compile.
Offline
I like the principle of the idea, it's just it might be hard to program a speech recognition interpreter, for scratch within scratch. I think it's a little to hard for the scratch team. It would be nice for an experiment though.
Offline
billyedward wrote:
Speech recognition is not impossible, but it is cumbersome and buggy.
Unless there is a good squeak library with those functions in it, I cannot see the scratch team investing their time in this.
If it is possible in a normal computer, then I don't see how they could do it anyway?After all, They could just use a better squeak library =, Or just use a whole different library, Other than Squeak. That library almost killed my laptop.
Offline
ThePCKid wrote:
gershmer wrote:
Richard2000 wrote:
Actually, It's possible on a Macintosh/Mac.And on Windows.
Only Windows Vista, Windows 7, and Macs have speech recognition.
And Google Translate.
(WinXP and 2000 can have speech recognition with copying the EXE of speech recognition to the other OS. )
Offline